Chris,
This approach works if we get it the right way round.
The second row should be Receiver Sensitivity
OMAouter, (min), so that the limit value never goes below -4.5 dBm for -FR1.
Dave
From: Chris Cole <chris.cole@xxxxxxxxxxx>
Sent: Wednesday, April 08, 2020 12:06 PM
To: David Lewis <David.Lewis@xxxxxxxxxxxx>; STDS-802-3-100G-OPTX@xxxxxxxxxxxxxxxxx
Subject: RE: [External]: [802.3_100G-OPTX] Comments #77 and #83
Hi Dave,
Yes, I understand the difference between transmitter and receiver tables. The subtle clue is the table title. So what difference does that make?
We have hockey stick shaped curve specs, for transmitter it’s OMA (min) vs. TDP (or OMA (min) vs. TDECQ), and for receiver it’s OMA (max) vs. SECQ. It’s been perfectly clear when numeric values have been used in the transmitter tables since
802.3ae. How is the receiver table so dramatically different that numeric values are not clear, and we need equation references? The equation references were a bad idea when introduced for OMA (max). That’s why we fixed this in January. So let’s not snatch
defeat from the jaws of victory as proposed by the Editor.
Let’s refine the January fix along the direction proposed by Peter Stassar, for example:
|
Description |
100GBASE-FR1 |
100GBASE-LR1 |
Unit |
|
Receiver Sensitivity in OMAouter minus SECQ (max) |
-5.9 |
-7.5 |
dBm |
|
Receiver Sensitivity OMAouter, (max)
|
-4.5 |
-6.1 |
dBm |
This approach has been crystal clear in 802.3 transmitter tables for 20 years. Its hard to argue that it will be confusing in 802.3 receiver tables.
As a bonus, we can get rid of the equations in the text because they become superfluous. And we can even make the Editor happy by getting rid of the table footnotes.
Then we can all relax in a hot tub wearing N95 masks, surgical gloves, and singing muffled Kumbaya.
Chris
From: David Lewis <David.Lewis@xxxxxxxxxxxx>
Sent: Wednesday, April 08, 2020 11:00 AM
To: Chris Cole <chris.cole@xxxxxxxxxxx>;
STDS-802-3-100G-OPTX@xxxxxxxxxxxxxxxxx
Subject: RE: [802.3_100G-OPTX] Comments #77 and #83
Chris,
The Tables you reference are to do with transmitters. The equations for receiver sensitivity were introduced in 802.3bs.
The same approach was followed in 802.3cd, 802.3cn and till now, in 802.3cu. So all of the SMF PAM4 receivers have taken advantage of the idea of “real receiver sensitivity” as measured with a practical transmitter. Previous clauses specified
a theoretical receiver sensitivity, measured with a perfect transmitter having no stress.
One problem with the above approach is that if we want receiver sensitivity to be normative we can’t put the requirements in a section dealing with parameter descriptions and test methods – they need to be in the tables and referenced by
the PICs. Peter’s proposed change is a neat way of achieving that without losing the requirement that measured receiver sensitivity needs to be below the limit for the value of SECQ used for the test.
Dave Lewis
From: Chris Cole <chris.cole@xxxxxxxxxxx>
Sent: Wednesday, April 08, 2020 9:13 AM
To: STDS-802-3-100G-OPTX@xxxxxxxxxxxxxxxxx
Subject: [802.3_100G-OPTX] Comments #77 and #83
Dear 802.3cu Task Force Participants,
Editor’s comments #77 and #83 argue that it is necessary to put back equation references into the specifications table to achieve clarity.
Proposal for Receiver Sensitivity (RS) Equation References in tables (comments
#77, #83)
This is fixing something that has worked perfectly well in 802.3 for the past two decades.
802.3ae TF introduced the concept of OMA (min) minus TDP to enable trade-off between optical TX power and TX penalty. The OMA (min) vs. TDP curve has a hockey stick shape, exactly like the OMA (max) vs. SECQ curve in 802.3cu. Let’s take
a look if the 802.3ae-2002 Standard found it necessary to insert equation references into the spec table to achieve clarity.
No equation references are to be found in the spec table, only numerical values.
Looking in 802.3ba-2010, 802.3bs-2017, 802.3cd-2018, we similarly find no equation references for OMA (min). Numerical values are perfectly clear.
Shockingly, even in 802.3cu, the same hockey stick shaped curve specifications for OMA (min) vs. TDECQ are numerical values, with no need for equation references in the spec table.
Yet in the same Standard, the OMA (max) vs. SECQ hockey stick shaped curves are deemed so different and so complicated that equation references in the spec tables are proposed as the only path to clarity.
A quote from the January presentation proposing the restoration of numerical values to the spec tables, and removal of equation references, is worth repeating.
http://www.ieee802.org/3/cu/public/Jan20/cole_3cu_01b_0120.pdf
If it ain’t broke, don’t fix it.
Thank you
Chris
From: Chris Cole
Sent: Tuesday, April 07, 2020 1:53 PM
To: Peter Stassar <Peter.Stassar@xxxxxxxxxx>;
STDS-802-3-100G-OPTX@xxxxxxxxxxxxxxxxx
Subject: RE: Your presentation on comments #77 and #84
The Editor must have missed the January presentation where the End Users were very clear on their preference.
http://www.ieee802.org/3/cu/public/Jan20/cole_3cu_01b_0120.pdf#page=29
It would sever the IEEE better if we wrote the specifications for them, rather than for our own notion of aesthetics.
Chris
From: Peter Stassar <Peter.Stassar@xxxxxxxxxx>
Sent: Tuesday, April 07, 2020 5:57 AM
To: STDS-802-3-100G-OPTX@xxxxxxxxxxxxxxxxx
Subject: [802.3_100G-OPTX] Your presentation on comments #77 and #84
Hi Gary and cu colleagues,
I just had a look at your presentation proposing responses to #77 and #84.
While I would agree that the changes agreed in Geneva could cause potential confusing, but at the same time I would agree with Chris, having proposed the changes during the Geneva meeting, that having just the
formula’s would also be confusing.
I believe there is another possibility for addressing the 2 related comments, which is by doing both, namely where we maintain the line in the Table for SECQ up to 1.4 dB with a fixed value and another line for
SECQ higher than 1.4 dB where we show the formula.
Then we don’t need the extensive note.
Furthermore we will also maintain the intent of the change made in D2.0.
There is precedence of dual line entrances in the various Transmitter Tables, where we distinguish between values for power above and below a certain extinction ratio.
Looking forward to our further discussion today.
Kind regards,
Peter
To unsubscribe from the STDS-802-3-100G-OPTX list, click the following link:
https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-100G-OPTX&A=1
To unsubscribe from the STDS-802-3-100G-OPTX list, click the following link:
https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-100G-OPTX&A=1
CAUTION: This email originated outside of Lumentum. DO NOT CLICK links or attachment unless you recognize the sender and know the content is SAFE.
To unsubscribe from the STDS-802-3-100G-OPTX list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-100G-OPTX&A=1